The opportunities and challenges of basket studies

There has been a growing emergence in the utilisation of basket studies and it’s not difficult to see why. Progress in genomics, tumour biology and statistics has led to advances in “precision oncology”. Cancers that were once viewed as homogeneous in terms of location and treatment strategy are now better understood to be increasingly heterogeneous across biomarkers and genetically determined subgroups. No two cancers are the same; tumours differ from patient to patient and few patients may noticeably improve with treatment, whilst others experience no benefit at all. As a result, we have seen a shift towards targeted agents, and it has become more common for trials to focus on a specific mutation at a particular location.

When thinking about the timeframe from initial drug discovery to regulatory review, we need to embrace novel clinical trial designs that improve efficiency. With this in mind, there has been an observable trend towards investigating multiple target-treatment pairs in parallel, either within, or across tumour types. The term “master protocol” refers to a general framework whereby multiple parallel drug studies are operated under one overarching protocol. A basket trial is a type of master protocol that tests the effect of one drug on a single mutation in a variety of tumour types. That is, they include patients with a certain genetic mutation in common regardless of the site or origin of cancer in the body, so patients have cancer at a variety of sites such as lung, breast, prostate, etc.

The major strength of basket studies is that they can be efficient in identifying treatments for cancers at several different sites with the same genetic mutation. If a drug is already approved for treating cancer at a particular location, a basket study can be used to see if the efficacy translates to other locations in the body. An example of this is vemurafenib (Zelboraf), which was originally approved by the FDA for the treatment of metastatic melanoma with BRAF V600 mutation. The
VE-BASKET study was an open-label, phase II, non-randomised, basket study investigating the use of vemurafenib for people with BRAF V600 mutation-positive cancers and other diseases. Final results showed that vemurafenib was also effective in treating a rare blood cancer known as Erdheim-Chester Disease (ECD), where patients have BRAF V600 genetic mutation. In November 2017, the FDA announced approval of vemurafenib for ECD with BRAF V600 mutation. That being said, the most common use for basket studies is still as early or mid-stage clinical research to help evaluate which potential indications for a drug would be good candidates for larger trials with more specific target in terms of location. In May 2017, however, the FDA granted approval to a treatment for patients whose cancers have a specific genetic feature, regardless of location, for the first time. Pembrolizumab (Keytruda) was indicated for the treatment of patients with solid tumours that are identified as having a biomarker referred to as microsatellite instability-high (MSI-H) or mismatch repair deficient (dMMR).

A further advantage of basket studies is that they only require one genetic assay for screening patients prior to study enrolment. Basket studies are also very flexible, allowing new baskets to be added and baskets showing futility to be dropped. And finally, cohorts within basket studies are often small and utilise single-stage or two-stage designs, which yield quick results, given sufficient patient numbers. Which brings us on to the potential limitations of basket studies. Some studies, or baskets within studies may have sample sizes that are too small, particularly if the genetic mutation of interest is rare.

Further limitations of basket studies are uncertainty around the assumption that genetic/molecular profiling may be sufficient to replace tumour histology, and the validity of designing a treatment that matches only a single mutation, when tumours may harbour multiple mutations at a time. Indeed, there may be more patients enrolled who only partially match the intervention (with multiple genetic mutations) than patients who fully match the intervention (with only one genetic mutation).

Basket studies may also not include a comparator arm, and when this occurs it can be difficult to interpret efficacy. Furthermore, tumour subtype is often prognostic, and there will also undoubtably be substantial heterogeneity in treatment efficacy across tumour types.

These numerous challenges in the study design need to be carefully considered as the number of basket studies being carried out continues to grow, but the associated strengths are too beneficial to ignore. The future is looking bright for basket studies and it will be interesting to see if they gain in popularity in the future, either for exploratory purposes, or as pivotal studies for FDA submission to gain approval for new drugs. Here at PHASTAR, we are very excited to be supporting the analysis and reporting of basket studies - read our case studies for more information.